In this paper, the mechanical faults of transformers including the winding radial deformation and axial displacement on 1.6燤VA transformer winding are investigated. Then, by estimating the parameters of the detailed model of this transformer winding in MATLAB software and changing these parameters in a manner that is proportional to the mechanical defects in electro-magnetic transients program software, the sampled differential current of the transformer is extracted for each disturbance. Next, the internal and external electrical faults and inrush current of the transformer are simulated. Afterwards, these signals are analyzed using maximal overlap discrete wavelet transform with Daubechies4 wavelet function, and their features are extracted. These extracted features are considered for training the classifiers of Decision Tree and artificial neural network. According to the simulation results, the proposed procedure is capable of classifying and discriminating among winding mechanical defects, internal and external electrical faults, and inrush current with a good accuracy that is the main novelty of this paper in comparison to other published works, which are limited to classifying only some of the mentioned faults.

中文翻译：

---

变压器绕组机械缺陷、内外电气故障及励磁涌流的分类与判别


本文对1.6燤VA变压器绕组的绕组径向变形和轴向位移等变压器机械故障进行了研究。然后，通过在 MATLAB 软件中估计该变压器绕组的详细模型的参数，并以与电磁瞬变程序软件中的机械缺陷成比例的方式改变这些参数，提取每个变压器的采样差动电流。干扰。接下来，对变压器的内部和外部电气故障以及励磁涌流进行仿真。然后，使用Daubechies4小波函数的最大重叠离散小波变换对这些信号进行分析，并提取它们的特征。这些提取的特征被考虑用于训练决策树和人工神经网络的分类器。根据仿真结果，所提出的程序能够以良好的精度对绕组机械缺陷、内部和外部电气故障以及浪涌电流进行分类和区分，这是本文与其他已发表的作品相比的主要新颖之处。仅限于仅对某些提到的故障进行分类。